SVNet: Where SO(3) Equivariance Meets Binarization on Point Cloud Representation

TL;DR

This paper introduces SVNet, a framework combining SO(3) equivariance and network binarization for 3D point cloud processing, enhancing efficiency and robustness in real-time applications like autonomous driving.

Contribution

It proposes a novel method to integrate scalar and vector features, enabling binarization while maintaining SO(3) equivariance in 3D neural networks.

Findings

Achieves a good balance between efficiency, robustness, and accuracy.

Demonstrates effectiveness on ModelNet40, ShapeNet, and ScanObjectNN datasets.

Applicable to backbones like PointNet and DGCNN.

Abstract

Efficiency and robustness are increasingly needed for applications on 3D point clouds, with the ubiquitous use of edge devices in scenarios like autonomous driving and robotics, which often demand real-time and reliable responses. The paper tackles the challenge by designing a general framework to construct 3D learning architectures with SO(3) equivariance and network binarization. However, a naive combination of equivariant networks and binarization either causes sub-optimal computational efficiency or geometric ambiguity. We propose to locate both scalar and vector features in our networks to avoid both cases. Precisely, the presence of scalar features makes the major part of the network binarizable, while vector features serve to retain rich structural information and ensure SO(3) equivariance. The proposed approach can be applied to general backbones like PointNet and DGCNN. Meanwhile, experiments on ModelNet40, ShapeNet, and the real-world dataset ScanObjectNN, demonstrated that the method achieves a great trade-off between efficiency, rotation robustness, and accuracy. The codes are available at https://github.com/zhuoinoulu/svnet.